This commit is contained in:
Dull Bananas 2024-05-21 19:38:11 +00:00
parent 6eb8d046fa
commit cf5afc9fb3
2 changed files with 367 additions and 231 deletions

View file

@ -5,10 +5,14 @@ use crate::schema::previously_run_sql;
use anyhow::{anyhow, Context};
use diesel::{
connection::SimpleConnection,
dsl::exists,
expression::IntoSql,
migration::{Migration, MigrationSource, MigrationVersion},
pg::Pg,
select,
sql_types,
update,
BoolExpressionMethods,
Connection,
ExpressionMethods,
NullableExpressionMethods,
@ -21,9 +25,15 @@ use lemmy_utils::{error::LemmyResult, settings::SETTINGS};
use std::time::Instant;
use tracing::info;
diesel::table! {
pg_namespace (nspname) {
nspname -> Text,
}
}
// In production, include migrations in the binary
#[cfg(not(debug_assertions))]
fn get_migration_source() -> diesel_migrations::EmbeddedMigrations {
fn migrations() -> diesel_migrations::EmbeddedMigrations {
// Using `const` here is required by the borrow checker
const MIGRATIONS: diesel_migrations::EmbeddedMigrations = diesel_migrations::embed_migrations!();
MIGRATIONS
@ -31,52 +41,75 @@ fn get_migration_source() -> diesel_migrations::EmbeddedMigrations {
// Avoid recompiling when migrations are changed
#[cfg(debug_assertions)]
fn get_migration_source() -> diesel_migrations::FileBasedMigrations {
fn migrations() -> diesel_migrations::FileBasedMigrations {
diesel_migrations::FileBasedMigrations::find_migrations_directory()
.expect("failed to find migrations dir")
.expect("failed to get migration source")
}
/// This SQL code sets up the `r` schema, which contains things that can be safely dropped and replaced
/// instead of being changed using migrations. It may not create or modify things outside of the `r` schema
/// (indicated by `r.` before the name), unless a comment says otherwise.
const REPLACEABLE_SCHEMA: &[&str] = &[
"CREATE SCHEMA r;",
include_str!("../replaceable_schema/utils.sql"),
include_str!("../replaceable_schema/triggers.sql"),
];
struct MigrationHarnessWrapper<'a, 'b> {
conn: &'a mut PgConnection,
options: &'b Options,
fn replaceable_schema() -> String {
[
"CREATE SCHEMA r;",
include_str!("../replaceable_schema/utils.sql"),
include_str!("../replaceable_schema/triggers.sql"),
]
.join("\n")
}
impl<'a, 'b> MigrationHarness<Pg> for MigrationHarnessWrapper<'a, 'b> {
const REPLACEABLE_SCHEMA_PATH: &str = "crates/db_schema/replaceable_schema";
struct MigrationHarnessWrapper<'a, 'b, 'c> {
conn: &'a mut PgConnection,
lock_conn: &'b mut PgConnection,
options: &'c Options,
}
impl<'a, 'b, 'c> MigrationHarnessWrapper<'a, 'b, 'c> {
fn run_migration_inner(
&mut self,
migration: &dyn Migration<Pg>,
) -> diesel::migration::Result<MigrationVersion<'static>> {
let start_time = Instant::now();
let result = rollback_if_lock_conn_broke(&mut self.conn, &mut self.lock_conn, |conn| {
conn.run_migration(migration)
});
let duration = start_time.elapsed().as_millis();
let name = migration.name();
info!("{duration}ms run {name}");
result
}
}
impl<'a, 'b, 'c> MigrationHarness<Pg> for MigrationHarnessWrapper<'a, 'b, 'c> {
fn run_migration(
&mut self,
migration: &dyn Migration<Pg>,
) -> diesel::migration::Result<MigrationVersion<'static>> {
let name = migration.name();
#[cfg(test)]
if self.options.enable_diff_check {
let before = diff_check::get_dump(&mut self.conn);
self.conn.run_migration(migration)?;
self.conn.revert_migration(migration)?;
let before = diff_check::get_dump();
self.run_migration_inner(migration)?;
self.revert_migration(migration)?;
let after = diff_check::get_dump();
diff_check::check_dump_diff(
&mut self.conn,
after,
before,
&format!("migrations/{name}/down.sql"),
&format!(
"These changes need to be applied in migrations/{}/down.sql:",
migration.name()
),
);
}
let start_time = Instant::now();
let result = self.conn.run_migration(migration);
let duration = start_time.elapsed().as_millis();
info!("{duration}ms run {name}");
result
self.run_migration_inner(migration)
}
fn revert_migration(
@ -84,12 +117,14 @@ impl<'a, 'b> MigrationHarness<Pg> for MigrationHarnessWrapper<'a, 'b> {
migration: &dyn Migration<Pg>,
) -> diesel::migration::Result<MigrationVersion<'static>> {
if self.options.enable_diff_check {
unimplemented!("diff check when reverting migrations");
//unimplemented!("diff check when reverting migrations");
}
let start_time = Instant::now();
let result = self.conn.revert_migration(migration);
let result = rollback_if_lock_conn_broke(&mut self.conn, &mut self.lock_conn, |conn| {
conn.revert_migration(migration)
});
let duration = start_time.elapsed().as_millis();
let name = migration.name();
@ -103,26 +138,24 @@ impl<'a, 'b> MigrationHarness<Pg> for MigrationHarnessWrapper<'a, 'b> {
}
}
// TODO: remove when diesel either adds MigrationSource impl for references or changes functions to take reference
#[derive(Clone, Copy)]
struct MigrationSourceRef<T>(
// If this was `&T`, then the derive macros would add `Clone` and `Copy` bounds for `T`
T,
);
impl<'a, T: MigrationSource<Pg>> MigrationSource<Pg> for MigrationSourceRef<&'a T> {
fn migrations(&self) -> diesel::migration::Result<Vec<Box<dyn Migration<Pg>>>> {
self.0.migrations()
}
}
#[derive(Default)]
pub struct Options {
enable_forbid_diesel_cli_trigger: bool,
enable_diff_check: bool,
revert: bool,
revert_amount: Option<u64>,
redo_after_revert: bool,
run: bool,
amount: Option<u64>,
}
impl Default for Options {
fn default() -> Self {
Options {
enable_forbid_diesel_cli_trigger: false,
enable_diff_check: false,
revert: false,
run: true,
amount: None,
}
}
}
impl Options {
@ -140,13 +173,16 @@ impl Options {
pub fn revert(mut self, amount: Option<u64>) -> Self {
self.revert = true;
self.revert_amount = amount;
self.run = false;
self.amount = amount;
self
}
pub fn redo(mut self, amount: Option<u64>) -> Self {
self.redo_after_revert = true;
self.revert(amount)
self.revert = true;
self.run = true;
self.amount = amount;
self
}
}
@ -155,32 +191,27 @@ pub fn run(options: Options) -> LemmyResult<()> {
let db_url = SETTINGS.get_database_url();
// Migrations don't support async connection, and this function doesn't need to be async
let mut conn =
PgConnection::establish(&db_url).with_context(|| "Error connecting to database")?;
let new_sql = REPLACEABLE_SCHEMA.join("\n");
let migration_source = get_migration_source();
let migration_source_ref = MigrationSourceRef(&migration_source);
//let mut conn = PgConnection::establish(&db_url).context("Error connecting to database")?;
let mut conn = PgConnection::establish(&db_url)?;
// If possible, skip locking the migrations table and recreating the "r" schema, so
// lemmy_server processes in a horizontally scaled setup can start without causing locks
if !(options.revert
|| conn
.has_pending_migration(migration_source_ref)
.map_err(|e| anyhow!("Couldn't check pending migrations: {e}"))?)
if !options.revert
&& options.run
&& options.amount.is_none()
&& !conn
.has_pending_migration(migrations())
.map_err(convert_err)?
//.map_err(|e| anyhow!("Couldn't check pending migrations: {e}"))?)
{
// The condition above implies that the migration that creates the previously_run_sql table was already run
let sql_unchanged: bool = select(
previously_run_sql::table
.select(previously_run_sql::content)
.single_value()
.assume_not_null()
.eq(&new_sql),
)
.get_result(&mut conn)?;
let sql_unchanged = exists(
previously_run_sql::table.filter(previously_run_sql::content.eq(replaceable_schema())),
);
if sql_unchanged {
let schema_exists = exists(pg_namespace::table.find("r"));
if select(sql_unchanged.and(schema_exists)).get_result(&mut conn)? {
return Ok(());
}
}
@ -190,103 +221,191 @@ pub fn run(options: Options) -> LemmyResult<()> {
conn.batch_execute("SET lemmy.enable_migrations TO 'on';")?;
}
// Running without transaction allows pg_dump to see results of migrations
// TODO never use 1 transaction
let run_in_transaction = !options.enable_diff_check;
// Repurpose the table created by `has_pending_migration` for locking, which
// blocks concurrent attempts to run migrations, but not normal use of the table
let transaction = |conn: &mut PgConnection| -> LemmyResult<()> {
let mut wrapper = MigrationHarnessWrapper {
conn,
options: &options,
};
// * Prevent other lemmy_server processes from running this transaction simultaneously by repurposing
// the table created by `MigrationHarness::pending_migrations` as a lock target (this doesn't block
// normal use of the table)
// * Drop `r` schema, so migrations don't need to be made to work both with and without things in
// it existing
// Using the same connection for both the lock and the migrations would require
// running all migrations in the same transaction, which would prevent:
// * Diff checker using pg_dump to see the effects of each migration
// * Migrations using an enum value added in a previous migration in the same transaction
// * Inspection of the database schema after only some migrations succeed
PgConnection::establish(&db_url)?.transaction(|lock_conn| -> LemmyResult<()> {
info!("Waiting for lock...");
let lock = if run_in_transaction {
"LOCK __diesel_schema_migrations IN SHARE UPDATE EXCLUSIVE MODE;"
} else {
""
};
wrapper
.conn
.batch_execute(&format!("{lock}DROP SCHEMA IF EXISTS r CASCADE;"))?;
lock_conn.batch_execute("LOCK __diesel_schema_migrations IN SHARE UPDATE EXCLUSIVE MODE;")?;
info!("Running Database migrations (This may take a long time)...");
(|| {
if options.revert {
if let Some(amount) = options.revert_amount {
for _ in 0..amount {
wrapper.revert_last_migration(migration_source_ref)?;
}
if options.redo_after_revert {
for _ in 0..amount {
wrapper.run_next_migration(migration_source_ref)?;
}
}
} else {
wrapper.revert_all_migrations(migration_source_ref)?;
if options.redo_after_revert {
wrapper.run_pending_migrations(migration_source_ref)?;
}
}
} else {
wrapper.run_pending_migrations(migration_source_ref)?;
}
diesel::migration::Result::Ok(())
})()
.map_err(|e| anyhow!("Couldn't run DB Migrations: {e}"))?;
// Drop `r` schema, so migrations don't need to be made to work both with and without things in
// it existing
revert_replaceable_schema(&mut conn, lock_conn)?;
// Run replaceable_schema if newest migration was applied
if !(options.revert && !options.redo_after_revert) {
run_selected_migrations(&mut conn, lock_conn, &options).map_err(convert_err)?;
// Only run replaceable_schema if newest migration was applied
if (options.run && options.amount.is_none()) || !conn.has_pending_migration(migrations()).map_err(convert_err)? {
#[cfg(test)]
if options.enable_diff_check {
let before = diff_check::get_dump(&mut wrapper.conn);
// todo move replaceable_schema dir path to let/const?
wrapper
.conn
.batch_execute(&new_sql)
.context("Couldn't run SQL files in crates/db_schema/replaceable_schema")?;
// todo move statement to const
wrapper
.conn
.batch_execute("DROP SCHEMA IF EXISTS r CASCADE;")?;
// todo use different first output line in this case
diff_check::check_dump_diff(&mut wrapper.conn, before, "replaceable_schema");
let before = diff_check::get_dump();
run_replaceable_schema(&mut conn, lock_conn)?;
revert_replaceable_schema(&mut conn, lock_conn)?;
let after = diff_check::get_dump();
diff_check::check_dump_diff(before, after, "The code in crates/db_schema/replaceable_schema incorrectly created or modified things outside of the `r` schema, causing these changes to be left behind after dropping the schema:");
}
wrapper
.conn
.batch_execute(&new_sql)
.context("Couldn't run SQL files in crates/db_schema/replaceable_schema")?;
let num_rows_updated = update(previously_run_sql::table)
.set(previously_run_sql::content.eq(new_sql))
.execute(wrapper.conn)?;
debug_assert_eq!(num_rows_updated, 1);
run_replaceable_schema(&mut conn, lock_conn)?;
}
Ok(())
};
if run_in_transaction {
conn.transaction(transaction)?;
} else {
transaction(&mut conn)?;
}
})?;
info!("Database migrations complete.");
Ok(())
}
fn run_replaceable_schema(
conn: &mut PgConnection,
lock_conn: &mut PgConnection,
) -> LemmyResult<()> {
rollback_if_lock_conn_broke(conn, lock_conn, |conn| {
conn
.batch_execute(&replaceable_schema())
.with_context(|| format!("Failed to run SQL files in {REPLACEABLE_SCHEMA_PATH}"))?;
let num_rows_updated = update(previously_run_sql::table)
.set(previously_run_sql::content.eq(replaceable_schema()))
.execute(conn)?;
debug_assert_eq!(num_rows_updated, 1);
Ok(())
})
}
fn revert_replaceable_schema(
conn: &mut PgConnection,
lock_conn: &mut PgConnection,
) -> LemmyResult<()> {
rollback_if_lock_conn_broke(conn, lock_conn, |conn| {
conn
.batch_execute("DROP SCHEMA IF EXISTS r CASCADE;")
.with_context(|| format!("Failed to revert SQL files in {REPLACEABLE_SCHEMA_PATH}"))?;
// Value in `previously_run_sql` table is not set here because the table might not exist
Ok(())
})
}
fn run_selected_migrations(
conn: &mut PgConnection,
lock_conn: &mut PgConnection,
options: &Options,
) -> diesel::migration::Result<()> {
let mut wrapper = MigrationHarnessWrapper {
conn,
lock_conn,
options,
};
/*let revert = (
options.revert,
MigrationHarnessWrapper::revert_last_migration as fn(_, _) -> _,
MigrationHarnessWrapper::revert_all_migrations as fn(_, _) -> _,
);
let run = (
options.run,
MigrationHarnessWrapper::run_next_migration,
MigrationHarnessWrapper::run_pending_migrations,
);
for (condition, run_one, run_all) in [revert, run] {
if condition {
if let Some(amount) = options.amount {
for _ in 0..amount {
run_one(&mut wrapper, migrations())?
}
} else {
run_all(&mut wrapper, migrations())?
}
}
}*/
if options.revert {
if let Some(amount) = options.amount {
for _ in 0..amount {
wrapper.revert_last_migration(migrations())?;
}
} else {
wrapper.revert_all_migrations(migrations())?;
}
}
if options.run {
if let Some(amount) = options.amount {
for _ in 0..amount {
wrapper.run_next_migration(migrations())?;
}
} else {
wrapper.run_pending_migrations(migrations())?;
}
}
/* } else {
wrapper.run_pending_migrations(migrations())?;
}
if let Some(amount) = options.revert_amount {
for _ in 0..amount {
wrapper.revert_last_migration(migrations())?;
}
if options.redo_after_revert {
for _ in 0..amount {
wrapper.run_next_migration(migrations())?;
}
}
} else {
wrapper.revert_all_migrations(migrations())?;
if options.redo_after_revert {
wrapper.run_pending_migrations(migrations())?;
}
}*/
Ok(())
}
/// Prevent changes from being committed after `lock_conn` unexpectedly closes
fn rollback_if_lock_conn_broke<T, E>(
conn: &mut PgConnection,
lock_conn: &mut PgConnection,
mut f: impl FnMut(&mut PgConnection) -> Result<T, E>,
) -> Result<T, E>
where
E: From<anyhow::Error> + From<diesel::result::Error>,
{
conn.transaction::<T, E, _>(|conn| {
let result = f(conn)?;
select(true.into_sql::<sql_types::Bool>())
.execute(lock_conn)
.context("Connection used for lock unexpectedly stopped working")?;
Ok(result)
})
}
/// Makes `diesel::migration::Result` work with `anyhow` and `LemmyError`
fn convert_err(
err: Box<dyn std::error::Error + Send + Sync>,
//) -> impl std::error::Error + Send + Sync + 'static {
) -> anyhow::Error {
anyhow::anyhow!(err)
}
#[cfg(test)]
mod tests {
use super::*;

View file

@ -1,4 +1,3 @@
use diesel::{PgConnection, RunQueryDsl};
use lemmy_utils::settings::SETTINGS;
use std::{
borrow::Cow,
@ -7,22 +6,22 @@ use std::{
process::{Command, Stdio},
};
diesel::sql_function! {
fn pg_export_snapshot() -> diesel::sql_types::Text;
}
// It's not possible to call `export_snapshot()` for each dump and run the dumps in parallel with the
// `--snapshot` flag. Don't waste your time!!!
pub fn get_dump(conn: &mut PgConnection) -> String {
/*// Required for pg_dump to see uncommitted changes from a different database connection
// The pg_dump command runs independently from `conn`, which means it can't see changes from
// an uncommitted transaction. NASA made each migration run in a separate transaction. When
// it was discovered that
let snapshot = diesel::select(pg_export_snapshot())
.get_result::<String>(conn)
.expect("pg_export_snapshot failed");
let snapshot_arg = format!("--snapshot={snapshot}");*/
pub fn get_dump() -> String {
let output = Command::new("pg_dump")
.args(["--schema-only", "--no-owner", "--no-privileges", "--no-comments", "--no-publications", "--no-security-labels", "--no-subscriptions", "--no-table-access-method", "--no-tablespaces"])
.args([
"--schema-only",
"--no-owner",
"--no-privileges",
"--no-comments",
"--no-publications",
"--no-security-labels",
"--no-subscriptions",
"--no-table-access-method",
"--no-tablespaces",
])
.env("DATABASE_URL", SETTINGS.get_database_url())
.stderr(Stdio::inherit())
.output()
@ -36,17 +35,19 @@ pub fn get_dump(conn: &mut PgConnection) -> String {
const PATTERN_LEN: usize = 19;
// TODO add unit test for output
pub fn check_dump_diff(conn: &mut PgConnection, mut before: String, name: &str) {
let mut after = get_dump(conn);
pub fn check_dump_diff(mut after: String, mut before: String, label: &str) {
if after == before {
return;
}
// Ignore timestamp differences by removing timestamps
for dump in [&mut before, &mut after] {
for index in 0.. {
let Some(byte)=dump.as_bytes().get(index) else{break};
if !byte.is_ascii_digit() {continue;}
let Some(byte) = dump.as_bytes().get(index) else {
break;
};
if !byte.is_ascii_digit() {
continue;
}
// Check for this pattern: 0000-00-00 00:00:00
let Some((
&[a0, a1, a2, a3, b0, a4, a5, b1, a6, a7, b2, a8, a9, b3, a10, a11, b4, a12, a13],
@ -54,16 +55,16 @@ pub fn check_dump_diff(conn: &mut PgConnection, mut before: String, name: &str)
)) = dump
.get(index..)
.and_then(|s| s.as_bytes().split_first_chunk::<PATTERN_LEN>())
else {
break;
else {
break;
};
if [a0, a1, a2, a3, a4, a5, a6, a7, a8, a9, a10, a11, a12, a13]
.into_iter()
.all(|byte| byte.is_ascii_digit())
.into_iter()
.all(|byte| byte.is_ascii_digit())
&& [b0, b1, b2, b3, b4] == *b"-- ::"
{
// Replace the part of the string that has the checked pattern and an optional fractional part
let len_after = if let Some((b'.', s)) = remaining.split_first() {
{
// Replace the part of the string that has the checked pattern and an optional fractional part
let len_after = if let Some((b'.', s)) = remaining.split_first() {
1 + s.iter().position(|c| !c.is_ascii_digit()).unwrap_or(0)
} else {
0
@ -79,69 +80,85 @@ pub fn check_dump_diff(conn: &mut PgConnection, mut before: String, name: &str)
}
let [before_chunks, after_chunks] =
[&before, &after].map(|dump| chunks(dump).collect::<BTreeSet<_>>());
let only_b = before_chunks.difference(&after_chunks).copied().map(process_chunk).collect::<BTreeSet<_>>();
let only_a = after_chunks.difference(&before_chunks).copied().map(process_chunk).collect::<BTreeSet<_>>();
[&before, &after].map(|dump| chunks(dump).collect::<BTreeSet<_>>());
let only_b = before_chunks
.difference(&after_chunks)
.copied()
.map(process_chunk)
.collect::<BTreeSet<_>>();
let only_a = after_chunks
.difference(&before_chunks)
.copied()
.map(process_chunk)
.collect::<BTreeSet<_>>();
// todo dont collect only_in_before?
let [mut only_in_before, mut only_in_after] = [
only_b.difference(&only_a),
only_a.difference(&only_b),
]
.map(|chunks| {
chunks
.map(|chunk| {
(
&**chunk,
// Used for ignoring formatting differences, especially indentation level, when
// determining which item in `only_in_before` corresponds to which item in `only_in_after`
chunk.replace([' ', '\t', '\r', '\n'], ""),
)
})
.collect::<Vec<_>>()
});
let [mut only_in_before, mut only_in_after] =
[only_b.difference(&only_a), only_a.difference(&only_b)].map(|chunks| {
chunks
.map(|chunk| {
(
&**chunk,
// Used for ignoring formatting differences, especially indentation level, when
// determining which item in `only_in_before` corresponds to which item in `only_in_after`
chunk.replace([' ', '\t', '\r', '\n'], ""),
)
})
.collect::<Vec<_>>()
});
if only_in_before.is_empty() && only_in_after.is_empty() {
return;
}
let after_has_more =
only_in_before.len() < only_in_after.len();
let after_has_more = only_in_before.len() < only_in_after.len();
// outer iterator in the loop below should not be the one with empty strings, otherwise the empty strings
// would be equally similar to any other chunk
let (chunks_gt, chunks_lt) = if after_has_more
{
only_in_before.resize_with(only_in_after.len(),Default::default);
let (chunks_gt, chunks_lt) = if after_has_more {
only_in_before.resize_with(only_in_after.len(), Default::default);
(&mut only_in_after, &only_in_before)
} else {
only_in_after.resize_with(only_in_before.len(),Default::default);
only_in_after.resize_with(only_in_before.len(), Default::default);
(&mut only_in_before, &only_in_after)
};
let mut output = format!("These changes need to be applied in {name}:");
let mut output = label.to_owned();
// todo rename variables
for (before_chunk, before_chunk_filtered) in chunks_lt {
let default = Default::default();
//panic!("{:?}",(before_chunk.clone(),chunks_lt.clone()));
let (most_similar_chunk_index, (most_similar_chunk, most_similar_chunk_filtered)) = chunks_gt
let (most_similar_chunk_index, (most_similar_chunk, _)) = chunks_gt
.iter()
.enumerate()
.max_by_key(|(_, (after_chunk, after_chunk_filtered))| {
if
after_chunk.split_once(|c:char|c.is_lowercase()).unwrap_or_default().0 !=
before_chunk.split_once(|c:char|c.is_lowercase()).unwrap_or_default().0 {0}else{
diff::chars(after_chunk_filtered, &before_chunk_filtered)
.into_iter()
.filter(|i| matches!(i, diff::Result::Both(c, _)
if after_chunk
.split_once(|c: char| c.is_lowercase())
.unwrap_or_default()
.0
!= before_chunk
.split_once(|c: char| c.is_lowercase())
.unwrap_or_default()
.0
{
0
} else {
diff::chars(after_chunk_filtered, &before_chunk_filtered)
.into_iter()
.filter(|i| {
matches!(i, diff::Result::Both(c, _)
// `is_lowercase` increases accuracy for some trigger function diffs
if c.is_lowercase() || c.is_numeric()))
.count()}
if c.is_lowercase() || c.is_numeric())
})
.count()
}
})
.unwrap_or((0,&default));
.unwrap_or((0, &default));
output.push('\n');
let lines = if !after_has_more{diff::lines(&before_chunk,most_similar_chunk)}else{
diff::lines(most_similar_chunk, &before_chunk)};
for line in lines
{
let lines = if !after_has_more {
diff::lines(&before_chunk, most_similar_chunk)
} else {
diff::lines(most_similar_chunk, &before_chunk)
};
for line in lines {
match line {
diff::Result::Left(s) => write!(&mut output, "\n- {s}"),
diff::Result::Right(s) => write!(&mut output, "\n+ {s}"),
@ -151,7 +168,8 @@ pub fn check_dump_diff(conn: &mut PgConnection, mut before: String, name: &str)
}
//write!(&mut output, "\n{most_similar_chunk_filtered}");
if !chunks_gt.is_empty() {
chunks_gt.swap_remove(most_similar_chunk_index);}
chunks_gt.swap_remove(most_similar_chunk_index);
}
}
// should have all been removed
assert_eq!(chunks_gt.len(), 0);
@ -177,19 +195,16 @@ fn process_chunk<'a>(result: &'a str) -> Cow<'a, str> {
Cow::Owned(lines.join("\n"))
} else if result.starts_with("CREATE VIEW") || result.starts_with("CREATE OR REPLACE VIEW") {
// Allow column order to change
let is_simple_select_statement = result
.lines()
.enumerate()
.all(|(i, mut line)| {
line = line.trim_start();
match (i, line.chars().next()) {
(0, Some('C')) => true, // create
(1, Some('S')) => true, // select
(_, Some('F')) if line.ends_with(';') => true, // from
(_, Some(c)) if c.is_lowercase() => true, // column name
_ => false
}
});
let is_simple_select_statement = result.lines().enumerate().all(|(i, mut line)| {
line = line.trim_start();
match (i, line.chars().next()) {
(0, Some('C')) => true, // create
(1, Some('S')) => true, // select
(_, Some('F')) if line.ends_with(';') => true, // from
(_, Some(c)) if c.is_lowercase() => true, // column name
_ => false,
}
});
if is_simple_select_statement {
let mut lines = result
.lines()
@ -205,7 +220,9 @@ fn process_chunk<'a>(result: &'a str) -> Cow<'a, str> {
(placement, line)
});
Cow::Owned(lines.join("\n"))
}else{Cow::Borrowed(result)}
} else {
Cow::Borrowed(result)
}
} else {
Cow::Borrowed(result)
}